Skip to main content
OSTI.GOVU.S. Department of Energy Office of Scientific and Technical Information
U.S. Department of Energy
Office of Scientific and Technical Information
 

Direct Prediction of Phonon Density of States With Euclidean Neural Networks

Journal Article · · Advanced Science
 [1];  [1];  [2];  [1];  [1];  [1];  [1];  [3];  [1];  [1]
  1. Massachusetts Inst. of Technology (MIT), Cambridge, MA (United States)
  2. Lawrence Berkeley National Lab. (LBNL), Berkeley, CA (United States); Lawrence Berkeley National Lab. (LBNL), Berkeley, CA (United States). Center for Advanced Mathematics for Energy Research Applications
  3. Argonne National Lab. (ANL), Lemont, IL (United States). Advanced Photon Source (APS)
+ Show Author Affiliations

Machine learning has demonstrated great power in materials design, discovery, and property prediction. However, despite the success of machine learning in predicting discrete properties, challenges remain for continuous property prediction. The challenge is aggravated in crystalline solids due to crystallographic symmetry considerations and data scarcity. Here, the direct prediction of phonon density-of-states (DOS) is demonstrated using only atomic species and positions as input. Euclidean neural networks are applied, which by construction are equivariant to 3D rotations, translations, and inversion and thereby capture full crystal symmetry, and achieve high-quality prediction using a small training set of ≈ 103 examples with over 64 atom types. The predictive model reproduces key features of experimental data and even generalizes to materials with unseen elements, and is naturally suited to efficiently predict alloy systems without additional computational cost. Furthermore, the potential of the network is demonstrated by predicting a broad number of high phononic specific heat capacity materials. The work indicates an efficient approach to explore materials’ phonon structure, and can further enable rapid screening for high-performance thermal storage materials and phonon-mediated superconductors.

Research Organization:
Argonne National Laboratory (ANL), Argonne, IL (United States); Lawrence Berkeley National Laboratory (LBNL), Berkeley, CA (United States)
Sponsoring Organization:
USDOE Office of Science (SC), Basic Energy Sciences (BES); National Science Foundation (NSF); USDOE
Grant/Contract Number:
AC02-06CH11357; SC0020148; 122374; AC02-05CH11231; DESC0020148
OSTI ID:
1771218
Alternate ID(s):
OSTI ID: 1782666; OSTI ID: 1785778; OSTI ID: 1814918
Journal Information:
Advanced Science, Vol. 8, Issue 7; ISSN 2198-3844
Publisher:
WileyCopyright Statement
Country of Publication:
United States
Language:
English

References (51)

Phonon density of states in R 2 CuO 4 and superconducting R 1.85 Ce 0.15 CuO 4 ( R =Nd,Pr) journal July 1993
Unsupervised Machine Learning and Band Topology journal June 2020
Accelerating materials property predictions using machine learning journal September 2013
Machine learning modeling of superconducting critical temperature journal June 2018
Lattice Dynamics of Gold journal October 1973
Phonons and related crystal properties from density-functional perturbation theory journal July 2001
Large phonon band gap in Sr Ti O 3 and the vibrational signatures of ferroelectricity in A Ti O 3 perovskites: First-principles lattice dynamics and inelastic neutron scattering journal April 2008
Generalized phonon density of states of the layer compounds Bi 2 Se 3 , Bi 2 Te 3 , Sb 2 Te 3 and Bi 2 (Te 0.5 Se 0.5 ) 3 , (Bi 0.5 Sb 0.5 ) 2 Te 3 journal July 1981
Controlling an organic synthesis robot with machine learning to search for new reactivity journal July 2018
Theory of Superconductivity journal December 1957
Energy-free machine learning force field for aluminum journal August 2017
Ab initio calculation of the thermodynamic properties and atomic temperature factors of SiO 2 α-quartz and stishovite journal April 1995
Predicting interfacial thermal resistance by machine learning journal May 2019
Commentary: The Materials Project: A materials genome approach to accelerating materials innovation journal July 2013
Recent advances and applications of machine learning in solid-state materials science journal August 2019
Predicting Thermal Properties of Crystals Using Machine Learning journal December 2019
High-Throughput Machine-Learning-Driven Synthesis of Full-Heusler Compounds journal October 2016
Molecular Dynamics with On-the-Fly Machine Learning of Quantum-Mechanical Forces journal March 2015
Five thermal energy grand challenges for decarbonization journal August 2020
Machine learning in materials informatics: recent applications and prospects journal December 2017
Rethinking phonons: The issue of disorder journal November 2017
Phonon density of states and anharmonicity of UO 2 journal March 2014
High-throughput density-functional perturbation theory phonons for inorganic materials journal May 2018
Phonon density of states and superconductivity in Nd 1.85 Ce 0.15 CuO 4 journal February 1991
Predicting the Band Gaps of Inorganic Solids by Machine Learning journal March 2018
Materials discovery and design using machine learning journal September 2017
Graph Networks as a Universal Machine Learning Framework for Molecules and Crystals journal April 2019
Finding symmetry breaking order parameters with Euclidean neural networks journal January 2021
Machine Learning Force Fields: Construction, Validation, and Outlook journal December 2016
Bandgap prediction by deep learning in configurationally hybridized graphene and boron nitride journal February 2019
Inelastic neutron scattering from zircon journal December 1997
Heat capacity of Mg3Sb2, Mg3Bi2, and their alloys at high temperature journal August 2018
Machine-learning-assisted materials discovery using failed experiments journal May 2016
Machine learning of accurate energy-conserving molecular force fields journal May 2017
Universal fragment descriptors for predicting properties of inorganic crystals journal June 2017
A general-purpose machine learning framework for predicting properties of inorganic materials journal August 2016
Accurate interatomic force fields via machine learning with covariant kernels journal June 2017
Hydrogen bonding, halogen bonding and lithium bonding: an atoms in molecules and natural bond orbital perspective towards conservation of total bond order, inter- and intra-molecular bonding journal January 2014
Detection of topological materials with machine learning journal June 2020
Automatic Chemical Design Using a Data-Driven Continuous Representation of Molecules journal January 2018
Electron-phonon interactions from first principles journal February 2017
Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable Prediction of Material Properties journal April 2018
Active learning of uniformly accurate interatomic potentials for materials simulation journal February 2019
Phonon dispersion relations, phase transitions, and thermodynamic properties of Zr Si O 4 : Inelastic neutron scattering experiments, shell model, and first-principles calculations journal March 2006
Vibrational Properties of Metastable Polymorph Structures by Machine Learning journal September 2018
Can machine learning identify the next high-temperature superconductor? Examining extrapolation performance for materials discovery journal January 2018
Python Materials Genomics (pymatgen): A robust, open-source python library for materials analysis journal February 2013
Mesenchymal stem cells and their conditioned medium can enhance the repair of uterine defects in a rat model journal March 2018
Hydrogen Transfer in Energetic Materials from ReaxFF and DFT Calculations journal April 2017
Gd$_{12}$Co$_{5.3}$Bi and Gd$_{12}$Co$_{5}$Bi, Crystalline Doppelgänger with Low Thermal Conductivities journalarticle January 2016
High-throughput Density-Functional Perturbation Theory phonons for inorganic materials [Supplemental Data] collection April 2018

Similar Records

Related Subjects

97 MATHEMATICS AND COMPUTING
density of states
machine learning
phonons